Casey Newton, writing at Platformer:

On Monday, OpenAI introduced what could be its most ambitious platform play to date. At the company’s developer day in San Francisco, CEO Sam Altman announced apps inside ChatGPT: a way to tag other services in conversations with the chatbot that allow you to accomplish a range of tasks directly inside the chatbot. 

In a series of demonstrations, software engineer Alexi Christakis showed what ChatGPT looks like after it has turned into a platform. He tagged in educational software company Coursera to help him study a subject; he tagged in Zillow to search for homes in Pittsburgh. In one extended demo, he described a poster he wanted, and Canva generated a series of options directly within the ChatGPT interface. He then used Canva to turn that poster into a slide deck, also within the chatbot. 

Starting today, developers can build these integrations using OpenAI’s software development kit. In addition to those above, services that will work with the feature at launch include Expedia, Figma, and Spotify. In the next few weeks, OpenAI said that they would be joined by Uber, DoorDash, OpenTable, and Target, among others. 

Eventually, OpenAI plans to add a directory that users can browse to find apps that have been optimized for ChatGPT. 

When I wrote about ChatGPT Agent back in July, I said the future of generative artificial intelligence was application programming interfaces via the model context protocol, a suite of interoperable tools that allow AI vendors to connect with each other’s products. I remain set on that idea and think Agent and tools like it aren’t headed anywhere, which is why OpenAI’s Monday announcements intrigued me so much. These integrations, which OpenAI calls “apps” developed through the ChatGPT software development kit, are virtually APIs that connect external tools to ChatGPT’s interface. They can be hailed by mentioning them in a chat, and when ChatGPT fetches data from an external tool, it uses MCP.

What this isn’t, however, is an operating system. Truthfully, I find AI companies to be too reliant on this phrase — not everything has to be an operating system, and neither should it be. These integrations are not apps, and by tweaking the terminology slightly, I think OpenAI can enjoy more success in the space. OpenAI already tried apps once in 2023, calling them “GPTs”: custom versions of ChatGPT with instructions and APIs to allow integration with third-party services. Today, GPTs are obsolete and don’t even use the latest, best models from OpenAI. The “GPT Store” was meant to be a paid marketplace where users could subscribe to these bespoke chatbots and use other services within ChatGPT, but that never transpired. This sounds familiar, doesn’t it?

By reframing the conversation around apps, as OpenAI did on Monday, it puts the onus on “app developers” to make integrations for ChatGPT. This is just how OpenAI rolls these days, and I find it both rude and anathema to the company’s name: OpenAI. Nothing about this system is “open” because it requires third parties to come to OpenAI to build apps and receive a small slice of the billions of dollars OpenAI plans to make one day. (The company currently hemorrhages money; it incurs a loss on every query sent to ChatGPT, paid subscriber or not.) Google was so successful in the early 2000s because it jibed well with the open web, promoting the sharing of ideas on the internet. OpenAI, contrary to its name, promotes the antithesis of that.

Whatever the strategy is, it seems to be working for OpenAI: Over 800 million people use ChatGPT regularly, a staggering number for a product only three years old. But it’s not working for nor aligned with the company’s stated mission: to build AI that benefits all of humanity. Currently, ChatGPT only benefits OpenAI’s plans for world domination and money-making, not even its investors or users. People are falling in love with ChatGPT and killing themselves based on its instructions. I haven’t suddenly become an AI doomer in the last few months, but rather, I’ve soured on OpenAI as a company. Ever since its loss of talent — Mira Murati, its chief technology officer, and Dr. Ilya Sutskever, its chief scientist, last year — OpenAI has solely been focused on corporate interests under the sole leadership of Sam Altman, its chief executive, who doesn’t care nor pretend to care about AI’s role in helping humanity.

Like it or not, the open web is necessary for any product to be successful. OpenAI’s faithful user base today can largely be attributed to ChatGPT’s web search capabilities, which have made it an excellent tool for all kinds of research, advice, and problem-solving. But if the company plans to erode its reliance on and trust in the open web, it might make a few bucks at the expense of doing good for society.